BitNotes Electronics Series
Garden of Light
Project #1 - Single vs. Two-Head solar lights to extend recharge effectiveness - 01/04/2018
As stated above, my garden is in the shade. In fact, there is a large magnolia tree who's canopy is above the two story roof level and has spread significantly.
This garden lies between two long North/South running buildings. Depending on the time of year I get a mostly rectangular shaft of light that progresses across
the garden going East or West. On the shortest days of the year, this means I only get full sun at any spot for 30 to 60 minutes.
Standard dollar store type garden solar lights don't seem to last very long in my garden. I've tried switching from rechargeable batteries to Alkaline for a slight
boost in voltage. Unfortunately, no matter what I try I can't get the batteries to hold enough charge with the existing circuitry to last more than about a week at best.
So that led me to the decision to try modifying the circuits in various ways and see what it takes to get a light that recovers during the day what power it has
consumed overnight. In some senses, winter is the perfect time to try this as this is the worst case scenario where the battery drains the most over the long nights and only
has just so long to recover during the day charging from the solar cells. If I can build something that works under these conditions, it should work pretty much all year long.
Construction here is fairly straightforward. In a nutshell this is two "heads" wired together in either parallel or series.
In the video I detail how to wire this project in series, but after a few days I got curious and built another one with solar cells wired in series.
For some unknown reason, after those few days, all three shut down relatively quickly in the evening, so I pulled them all, recharged the batteries and put them back out angled a bit more toward the sun.
For the uninitiated, parallel wiring means that all negative wire are wired together and all positives are wired together.
Series is like a chain. the positive of one cell is wired to the negative of the next cell. For only two solar cells this means only one connection in the middle and then the two outside wires
are run back to the circuit. The result is that for parallel you are combining the current of both solar cells without increasing voltage. For series, you are combining the voltage of both cells.
To run the wires between the two heads, I drilled to holes large enough to fit the wires and threaded them through.
I chose one head to light up and removed the circuit board and battery contacts from the other leaving only the wires coming from that solar cell.
With that, there is more space in the head witout the circuit board or battery, so if you need more space for connections that's a good place to put them.
With all the wiring done, there is still the issue of how to mount both heads. I simply joined them together side by side with a bit of hot melt glue.
I learned after moving the construction a few times that that configuration is best served by a significant amount of hot melt glue.
Other ways to join them include super-glue or physical fasteners like small nuts and bolts.
Another possible configuration could be to use longer wires between the cells and mounting the second head some distance away from the first.
With that you could angle the second head differently to account for different angles of the sun at different times of the year.
If you're situation is similar to mine, with a shaft of sun moving across the garden, you could move the second head far enough away to extend the amount of time each cell is exposed to the sun.
There really are a number of possibilities on how you could set these two heads up. You could even daisy chain them across your garden to power just the one light.
Data (This is where most of the updates over time will be)
First of all, after some earlier testing, which I'll detail in an upcoming video on recharging batteries, I learned that the NiMh cells I have have a capacity of 60 mAH.
That number comes from fully charging multiple batteries and then discharging them down to 1 Volt and measuring the time it took to reach that voltage.
Also, I tested the draw from the LEDs while connected to the solar lamp circuitry.
Since both the green and white LEDs drew 1.86 and 2.00 mA and the main chip in the circuit is designed to limit current based on the value of the inductor, I'm concluding that the limit is set to 2mA.
So straight out of the gate, the calculated amount of time the LED could be run continuously would be 60 mAH / 2mA = 30 Hours.
That's more than enough for one night, so each of the rigs I built should be on from sundown to sunrise the first couple of nights.
From there it's a question of how much power (mAH) the solar cells and circuit recharge the battery during the day.
There may be theoretical numbers for the charge during daylight hours, but those are affected by, amongst other things, the "shadiness" of the garden and whether or not the sky is overcast.
So the best I can do is just measure the hours each rig is on each night and compare the original solar garden light to the two new rigs, one wired in series and one in parallel.
Each is located in the same area of the garden and within view of a security camera that happens to record both motion activated video AND timelapse with one frame ever second.
I should mention that it occured to me early on to measure the voltage of the battery every day at the same time, but I realized that wouldn't be reliable.
While Alkaline batteries tend to discharge linearly, rechargeable cells tend to drop to a certain voltage and hold there until almost all available power is gone.
So while it might be that the batteries have the same voltage when checked each day, they could easily have less and less charge each day and that wouldn't be apparent until the voltage falls off
after a number of days.
A better, though not perfect, way to measure available power is to track when the circuit shuts down the LED overnight and that's why I have the rigs set in the garden in view of my camera.
The downside is that to do that I had to place the 3 rigs in one of the darkest corners of the garden because the camera wasn't intended to watch the garden.
It just happens to capture that corner of the garden.
The following table shows each rig and the data I've collected so far. Unfortunately, something happened and all three rigs suddenly started shutting off after a very short period of time.
One even turned on and off a couple of times and then just didn't come back on.
My best guess was a loose connection around the switch that allows you to turn the whole rig off.
To eliminate that variable, I have not soldered the contacts into the "on" position.
I then recharged all the batteries and set them back out in the same location but angled a bit more towards the sky visible between the buildings.
|Day / Night
||Twin Head Parallel
||Single Head (Control)
||Twin Head Series*
Two-Head rig wired in series was installed during third cycle and subsequently changed (see below)
Two-Head rig wired in parallel only stayed on for 2 minutes! It actually came on then off then on then off again within those two minutes. Something seems to be wrong.
Control only lasted for 25 minutes.
After the third cycle there were a few issues. My best guess was that the switch might be causing the problems, though it would be unusual that it would happen to both active rigs at the same time.
Also both two-head rigs had failures at the hot glue joint.
I chose to remove all three rigs, recharge all batteries with my dumb charger and resecure the hot glue connections.
I then placed all three rigs back in the garden (still in view of the camera) in roughly the same location at 2:10 PM so they had a minor chance to top off their respective charges.
I also angled them a bit better to hopefully catch a bit more sun.
As this page goes up they have only just turned back on and I'll add another tracking table after sunrise.
As of this writing, January 4 2018, there just isn't enough data to explore further tweaks.
My suspicion though is that I may have to either go to three heads or change out the inductor in the circuit which controls how bright the light is and how much current it uses.
In case anyone was wondering, I did say I built two of the parallel wired two-head rigs. The other is sitting in the main part of the garden and has been running for about a week.
That's the one with the green LED and I haven't collected any numerical data.
However, I can say that it seems to last longer than the ones in camera view while still turning off earlier and earlier each night.
I does seem to recover just enough charge to stay on for at least a couple hours each night, but my goal is to use this experiment to determine what it would take to build a rig
that would last through the night. I want a rig that will shut off when the sun comes up, not when the battery runs out.
Granted, I'm doing this experiment when the nights are very long, but that could be the best time to do it. If it can last through the night in the peak of winter, it should be
able to last all year.